Materials Informatics with PoreBlazer v4.0 and the CSD MOF Database

The development of computational methods to explore crystalline materials has received significant attention in the last decades. Different codes have been reported to help researchers to evaluate and learn about the structure of materials and to understand and predict their properties. In this Methods article, we present an updated version of PoreBlazer, an open-access, open-source Fortran 90 code to calculate structural properties of porous materials. The article describes the properties calculated by the code, their physical meaning, and their relationship to the properties that can be measured experimentally. Here, we reflect on the methods in the code and discuss features of the most recent version. First, we demonstrate the capabilities of PoreBlazer on the prototypical metal-organic framework (MOF) materials, HKUST-1, IRMOF-1, and ZIF-8, and compare the results to those obtained with other codes, Zeo++ and RASPA. Second, we apply PoreBlazer to the recently assembled database of MOF materials-the CSD MOF subset-and compare properties such as the accessible surface area and pore volume from PoreBlazer and the two other codes, and reflect on the possible sources of the differences. Finally, we use PoreBlazer to illustrate how correlations between various structural characteristics can be mined using interactive, dynamic data visualization and how material informatics approaches-including principal component analysis and machine learning-can accelerate the discovery of new materials and new functionalities. The results of these calculations, along with the PoreBlazer code, documentation, and case studies, are available online from https://github.com/SarkisovGroup/PoreBlazer. The data visualization tool is available at https://github.comi aarni-analytiestmotlexplorer, and the principal component analysis is available at https://github.com aaml-analytics/pca-explorer.